The Remote Associates Test: Science of Creative Connections

The Remote Associates Test (RAT) is a 40-item word puzzle that has been used in creativity research for over 60 years. Each item presents three seemingly unrelated words. Your job is to find a single fourth word that forms a compound word or common phrase with each of them.

Example: FENCE | CARD | MASTER Answer: POST — fence post, postcard, postmaster.

What makes the RAT interesting isn't just that it's a puzzle. Performance on it predicts creative output better than most other short cognitive assessments, and the neural mechanisms behind success on the test directly map to the mechanisms behind real-world creative insight.

How the Remote Associates Test Works

The test was developed by Sarnoff Mednick at the University of Michigan and published in 1962. Mednick designed it based on his associative theory of creativity: that creative thinking consists of forming associative elements into new combinations, and that the ability to make remote associations — connections between concepts that are far apart in semantic space — distinguishes highly creative thinkers from others.

Each RAT item requires you to find the one word in your vocabulary that has a strong enough association with three disparate words to link them all. The words are chosen to have no obvious shared theme, which forces semantic search well beyond your nearest associations.

A few more examples to make the structure concrete:

• PINE | CRAB | SAUCE → APPLE (pineapple, crab apple, applesauce)
• BLUE | CAKE | COTTAGE → CHEESE (blue cheese, cheesecake, cottage cheese)
• DREAM | BREAK | LIGHT → DAY (daydream, daybreak, daylight)

If you solved those quickly, your associative network activated the connecting word before you consciously searched for it. If you needed to work through them methodically, you're using a different strategy — and that difference is meaningful.

What the RAT Actually Measures

Mednick argued that creative people have "flatter" associative hierarchies. For most people, associations to any given word cluster tightly around the most common meanings. A flat hierarchy means more concepts are activated with roughly equal strength, so remote connections become accessible without deliberate effort.

For a word like "FIRE," a steep hierarchy activates: hot, flame, burn. A flat hierarchy also activates: truck, station, works, side, alarm, fly, place — with enough residual strength to surface when needed.

EEG studies by Mark Jung-Beeman and colleagues at Northwestern (2004) captured the neural signature of RAT solutions in real time. Insight solutions — the "aha" moment where the answer seems to arrive fully formed — were preceded by a burst of gamma-band activity in the right anterior temporal lobe, about 300 milliseconds before the answer reached consciousness. Analytical solutions, where people worked through candidates systematically, showed no such burst.

This gamma spike arises in a specific cognitive state: diffuse attention, low visual input, relative quiet. It represents the brain making a long-range semantic connection across cortical areas that focused, analytical attention tends to suppress. This is why the shower is actually a good place to solve RAT items — you're not suppressing the remote associations that focused effort would crowd out.

Why RAT Performance Predicts Creative Output

A meta-analysis by Benedek and colleagues (2012) found meaningful correlations between RAT performance and multiple measures of creative output, including divergent thinking scores and rated creative achievement. The test has also been used as the outcome measure in studies of:

• How positive mood states affect creative thinking (positive affect consistently improves RAT performance by loosening cognitive categories)
• How incubation periods help solve previously stuck problems
• Whether mindfulness meditation affects the frequency of creative insight
• The effects of very low alcohol doses on associative range

The convergence of these findings suggests the RAT is measuring something real about a general creative capacity, not just puzzle-solving skill. It's one of the few short creativity assessments that holds up well against real-world creative output measures.

RAT vs. Divergent Thinking Tests

The RAT and divergent thinking tests like the Alternative Uses Task measure different aspects of creativity. Divergent thinking tests ask you to generate many different answers — breadth across categories. The RAT requires convergence on a single correct answer — precision of semantic reach.

Neither predicts creative performance well on its own. High creativity seems to require both: the ability to generate multiple remote associations AND the ability to evaluate which connection is structurally valid.

This maps directly onto the relationship between divergent and convergent thinking. Divergent thinking opens possibilities; convergent thinking selects among them. The RAT essentially requires both operations in sequence: a wide semantic search followed by evaluation of which candidate actually satisfies all three cues simultaneously. It compresses the full creative cycle into a single brief task.

How to Build Associative Flexibility

Performance on the RAT improves with practice on tasks that train broad semantic activation.

Read across domains. Wide reading builds a denser, more interconnected semantic network. The more domains you have represented in memory, the more bridging concepts you have available when faced with remote association problems.

Use word association as a daily habit. Take a random word and spend two minutes listing associations, then associations of those associations. The goal isn't fluency — it's range. You want to reach the second and third rings of association, not just the obvious neighbors.

Practice divergent thinking. The alternative uses task trains the same categorical loosening the RAT requires. The divergent thinking exercise builds this directly and measures how many distinct categories your answers span.

Work the RAT itself. The Remote Associates exercise gives you RAT-format problems with immediate feedback. Because the answer is either right or wrong, it provides the kind of precise feedback that builds the skill efficiently — unlike open-ended creative tasks where quality is hard to measure.

Cultivate positive affect before creative work. Alice Isen's lab at Cornell demonstrated repeatedly through the 1980s and 1990s that mild positive affect improves RAT performance. The mechanism is reduced cognitive inhibition and broader semantic activation. This isn't an argument for forced positivity — it's a practical note about working conditions.

A Newer, Computational Version of the Same Idea

Mednick's associative theory predicted that semantic distance between concepts should be measurable and should correlate with creative ability. For decades that was hard to test directly — "distance" between words was an intuition, not a number. In 2021, Olson, Nahas, Chmoulevitch, Cropper, and Webb published a study in PNAS introducing the Divergent Association Task (DAT): participants list 10 nouns as unrelated to each other as possible, and a computer scores the semantic distance between every pair using word-embedding models trained on large text corpora.

The result validated Mednick's core claim with a precision the original 1962 test couldn't achieve. DAT scores — a pure measure of how far apart your generated concepts sit in computational semantic space — correlated with RAT performance, divergent thinking scores, and self-reported creative achievement, despite requiring no puzzle-solving skill and taking about 90 seconds to complete. It's evidence that the "flat associative hierarchy" Mednick proposed in 1962 is a real, computationally measurable property of how creative thinkers organize concepts, not just a theoretical construct built to explain RAT scores after the fact.

Limitations of the RAT

The test isn't a complete picture of creativity. Mednick's associative theory has been critiqued for overemphasizing verbal associations at the expense of visual, spatial, and embodied forms of creative thinking. Real breakthroughs in art, music, and engineering involve processes the RAT only partially captures.

The RAT is also susceptible to vocabulary effects. Native English speakers with large vocabularies have a structural advantage that reflects language exposure, not just associative flexibility. Cross-cultural comparisons require adapted versions.

What the RAT does well — measuring a specific kind of semantic flexibility with precise, binary feedback — is rare among creativity assessments. Most creative tasks don't give you a clear right or wrong answer. The RAT does. That clarity makes it uniquely useful as a training tool.

The connection to creative block research is also direct. Cognitive fixation — the mechanism underlying creative block — is essentially over-activation of a local semantic cluster at the expense of remote connections. Building remote associative flexibility through RAT practice makes you more resistant to that kind of fixation, not just better at word puzzles.

For a broader, open-ended assessment that measures generation rather than convergence, see the Torrance Tests of Creative Thinking — together, the two instruments cover most of what psychometric creativity research actually measures.

---

Ready to train your creativity? Try science-backed exercises that measure and improve your creative thinking. Start a Free Exercise